Designing user-centered workspaces starts with a mindset, not a room.


time-to-reclaim-the-open-plan-office-1

Time-to-reclaim-the-open-plan-office (1)

It’s Time to Reclaim the Open Office

The science is in: open offices have failed. Despite the utopian visions that led to their inception, real-world open offices—those fishbowls and theaters of distraction—have proven only to reduce efficiency and increase stress, as their grid-like layouts reinforce our interchangeability. Yet we keep designing and building them—frog included—because the economics and collaboration potential remain compelling, if elusive, design goals. The grand opening of an open-office innovation lab at <a href="https://www.frog.co/work/equinor">Equinor </a>this June was a proud moment for our <a href="https://www.frog.co/services/architecture-places" target="_blank" rel="noopener">Places</a> practice because it was the manifestation of months of intense thinking about how to succeed where so many others have failed.
The constraints our team faced were precisely those behind the plague of cube farms: a very limited space (2,120 sq. feet) that needed to accommodate the work of as many as three R&amp;D teams. Our biggest fear was that space constraints would constrain our thinking, just as economic optimizations degraded the original 1950s open office concept, Bürolandschaft <a href="https://en.wikipedia.org/wiki/Office_landscape">office landscape</a>, which broke the strict grid of offices and hierarchy into organic layouts and furniture systems focused on the individual, to the drab cubicle farms of the 1970s. These “veal-fattening pens” (Douglas Coupland’s term) borrowed only the physical form of an officeless layout and none of the emotive or aspirational qualities of the original open office concept.
If anything, the desire for adaptability has only increased during the digital age, thanks to multiplying modes of working that have replaced the 40-hour desk job. We therefore asked whether we could reinterpret the spirit of the “office landscape” for a modern workforce. But, you may ask: Is this wishful thinking? Just another case of designers falling in love with principles divorced from reality? We think not. In talking to employees we found that organic, flexible environments are even more relevant today. By examining the many dimensions of technology research within Equinor, we arrived at a multi-layered environment that adapts to the workers and the work being done, rather than requiring conformity of the workers to the spatial design.
Such multi-layered designs, drawing on the expertise of a wide range of disciplines, are becoming more common as the digital and physical worlds integrate. For the Equinor innovation lab, we brought together architecture, visual and interaction design, technology, business process, and user research to create a truly orchestrated environment. We were engaging in a form of <a href="https://uxplanet.org/can-speculative-design-make-ux-better-design-trend-4-4-ce8d13148e5d">speculative design</a>—the use of design methods in crafting a future. Each company faces many possible futures, some more humane for its workforce than others. For Equinor, we began with rigorous design research to gather individual needs and unpack workflows, determined technology requirements, then took a multi-layered approach to creating a space to serve these needs in the years to come.
For a physical space, user-centered design starts with the mindset, not a room. One of the main goals of design research was unpacking what the term ‘innovation’ meant for Equinor. We learned that for Equinor, innovation is about rapidly assembling global, cross-disciplinary teams to solve hard problems. For them, innovation means that everyone can contribute their expertise directly to a project, using the tools and methods of their discipline.
In other words, the workspace is the stage on which innovation is improvised, adapting to individual participants. The workplace needs to offer neighborhood spaces, team spaces, heads-down space, make-lab space, and event space. Unlocking the right mix of spaces was the essence of the design challenge. Every organization presents a unique blend of activities, requiring a unique mix of spaces. Getting this balance right allows users to select the environment they want to work in. Given Equinor’s small footprint, as well as the dynamic nature of an innovation lab, the design needed to adapt to accommodate all these spaces.
Since technology alone does not create a responsive environment, we needed to reconceive how we design interior spaces for easy reconfiguration of the layout and atmosphere. We designed components systemically to allow a space to be reprogrammed without a large investment from the organization or disruption to the team’s workflow. In Equinor’s case, the teams can switch up the space within a few minutes. The furniture is modular and on casters, making moving it to a new configuration or into one of the storage spaces a breeze. The technology (e.g., screens, cameras, power) was strategically placed to be usable across the range of spatial reconfigurations. The few permanently designed spaces were mainly for heads-down time and were located along the periphery.
Finally, with multiple teams working in one room, sound abatement was key. The ability to cancel out noise is critical to the functionality of the space and the focus of the teams. We solved this through a mix of sound-masking technology along with material and furniture selection.
The investment to renovate or build new is a massive undertaking for any organization. As designers, we have a duty to build environments that work, live and change with the organizations they are built for.
When Robert Propst of Hermann Miller designed the first cubicles, <a href="https://en.wikipedia.org/wiki/Action_Office">Action Office</a>, and then Action Office II, the emphasis was on customization: wall placement, personalization of private space, easy reconfiguration—a complete system for the <a href="https://www.wired.com/2014/04/how-offices-accidentally-became-hellish-cubicle-farms/">adaptable office of the 1960’s</a>. The error companies have made is in adopting the form and not the spirit of such systems, by applying configurations top-down and across the board to maximize one metric, the utilization of space, rather than letting the work—or “action—of a company define the space that supports it.
Now is our opportunity to pivot our discipline back toward the people who live and work in the places we create.

Elizabeth Wehr

Sheldon Pacotti

Elizabeth is an experience designer at frog in Austin. With a background as a licensed architect, her focus is on the integration of digital and physical experiences in the design of a place for people. Elizabeth believes that digital and physical can no longer be thought of as two separate and disparate entities, but rather as one in creating holistic experiences that are enabled by technology, but not focused on technology.

Sheldon is Senior Solution Architect at frog in Austin. Having studied math and English at MIT and Harvard, Sheldon enjoys cross-disciplinary creative projects. He builds award-winning software, writes screenplays for video games, creates software architectures for businesses and writes about technology

Authors

pacotti_headshot3_flipped_bw_square-300x300

So much of what is written about the future of Al amounts to broad philosophical statements, often about the nature of humanity. knowledge and life itself.

the_current_state_of_intelligent_systems_2-3

The Current State of Intelligent Systems

Avia, an <a style="font-family: 'BentonSans', sans-serif;" href="https://futurism.com/a-new-ai-can-write-music-as-well-as-a-human-composer/" target="_blank" rel="noopener">AI-powered music composer</a>, suggests algorithms are capable of remarkable creativity. <a style="font-family: 'BentonSans', sans-serif;" href="http://www.kurzweilai.net/googles-new-multilingual-neural-machine-translation-system-can-translate-between-language-pairs-even-though-it-has-never-been-taught-to-do-so" target="_blank" rel="noopener">Google Translate’s invention of its own interlingua</a>—the ability to translate between two languages it has not yet been trained to interpret—may lead some to fear the machines are on the brink of autonomy.
We asked Sheldon Pacotti, Senior Solution Architect at frog Austin, to unpack some of this complexity for us. Here he shares his view on the current state of AI in product design, its common misconceptions and a look at where we’re going next.
Q: Why does AI spark such philosophical debate?
A: What’s unfortunate—but fun—about AI is that it’s in fact deeply entwined with the field of philosophy. Analytical philosophy, from Hume to 20th century figures like <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Donald_Davidson_(philosopher)" target="_blank" rel="noopener">Donald Davidson</a>, has had a direct impact on how truth is represented in “symbolist” AI systems, for instance. As a computer scientist, I would say that W. V. Quine’s <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Willard_Van_Orman_Quine" target="_blank" rel="noopener">theory of “observation sentences”</a> and Jerry Fodor’s <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Jerry_Fodor" target="_blank" rel="noopener">Language of Thought</a> both read like software architecture.
It’s hard to imagine creating intelligent human-machine interfaces without considering epistemology, phenomenology and other -ologies developed by abstract thinkers. Complicating things further, a half-dozen other fields, including neuroscience, mathematics and cognitive science, are also core to the advance of AI. This situation makes it much more difficult to predict when we will devise the “right” theories of intelligence, and in what order these theories might appear. To build a thinking machine, you need a good theory of thought, perception and reasoning.
In design, through the practice of cultivating empathy, we routinely apply this kind of thinking to users. What is changing is that we must now consider how our creations perceive us, since they will be participating in our lives, communities and society at large. This quickly slides into philosophy, but I would say that we want empathy to be an internal organizing principle of any thinking system we design.
Q: It’s impossible to have a conversation about AI without also discussing the role of the data underlying the intelligence. Why does a “deep learning” system need so much data?
A: Deep learning systems run on data that has been collected, curated and engineered in a deliberate way to solve a specific problem. The large quantity of data feeds statistical learning that is very granular—down to the individual pixel for an image-recognition system, for instance. Just as a study involving statistics requires a large sample size to be accurate, “deep” neural nets require an extreme amount of training data in order to learn to decipher features with accuracy.
Q: We hear a lot about the importance of training the data in an intelligent system, which sounds like a repetitive and somewhat tedious process. If you need a million pictures in order to learn what a cat looks like, doesn’t that make you a slow learner?
A: Deep learning, inspired by the convolutional layers of the human neocortex, is just one aspect of cognition. People can learn new ideas so quickly because the human brain contains many other structures that contribute to intelligence, including “thinking neurons&#8221; believed to bind themselves deterministically to a single concept. These other structures enable what AI researchers call “<a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Transfer_learning" target="_blank" rel="noopener">transfer learning</a>,” the ability to leverage existing knowledge to quickly learn new concepts. As AI architectures advance, they will acquire some of these features and gain the ability to learn with less data.
At the other extreme from deep learning is the “symbolist” approach to AI, which uses formal logic to create massive ontologies (e.g. <a style="font-family: 'BentonSans', sans-serif;" href="https://www.w3.org/OWL/" target="_blank" rel="noopener">OWL</a>  and <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Cyc" target="_blank" rel="noopener">Cyc</a>) and smart expert systems capable of human-like deduction. Though the symbolist approach is out of fashion and even labeled the “old” approach to AI by many, it reflects a distinct quality of human thought not captured by the <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Convolutional_neural_network" target="_blank" rel="noopener">generic convolutional neural net</a>. The creation of systems able to learn from sparse data is likely to involve the merging of these two paradigms.
Q: Tell us a bit about the “transparency challenge.” How important is transparency to AI development from a technology standpoint?
A: The &#8220;<a style="font-family: 'BentonSans', sans-serif;" href="https://www.computerworld.com/article/3155534/artificial-intelligence/artificial-intelligence-transparency-isn-t-just-a-trend.html" target="_blank" rel="noopener">transparency challenge</a>” in AI is an especially difficult problem for mathematically trained systems like neural nets. What deep learning systems provide today are single-function black-box operations. For example, a system may be created to recognize a face from a photograph, or to translate a sentence. When they are integrated into a traditional software system, such as a strategy game, they take on a logical, discernable role, but their inner workings remain opaque. Looking ahead to neuromorphic computers, we can imagine architectures in which all of the steps in a system are learned, including the algorithms for executing these steps. The result could be a new class of spookily intelligent black boxes.
The key to making these systems “transparent” may lie in their working memory—a system analogous to the previously mentioned strategy-game AI. We will be engineering these systems at a high level, defining operations for storing patterns, focusing attention on patterns and so on, so we will have the hooks to trace the flow of execution. We will be able to isolate individual learned operations much as the human mind associates concepts with words. Sub-systems, like vision, might remain opaque, but at the pattern or concept level, we will learn to engineer transparency and even introspection.
Q: Bots seem to be everywhere now, but how much do they really understand? Are they all they’re cracked up to be?
A: Companies keep adding features to their bots, like they would to any app, but we seem to be a long way away from having a natural conversation with a computer. Though these systems make clever inferences using a variety of methods and offer a natural way to perform certain tasks, they aren’t really a part of our world. They don’t have mental models of what it’s like to live a human life. Some might say that this understanding, too, will drop out of big data analytics.
However, many leaders in the AI field believe that true artificial intelligence will need to be based on <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Embodied_cognition" target="_blank" rel="noopener">embodied intelligence</a>. This is the school of thought that a thinking machine needs to be physically sensing and even navigating the world. Though senses may seem like overkill for bots that help us shop, they could provide a unifying “format” for learned concepts, as they do in the human mind.
Q: What will be the next AI breakthrough? 
A: We are likely to see extended periods of “me-too” applications, while the latest systems learn all they can learn, followed by surprising leaps. Today’s deep learning is very good about “understanding&#8221; patterns in large datasets, given a narrow problem-area. New neuromorphic designs are on the horizon, however. Spearheaded by researchers at DeepMind, complete computers have been built entirely with neural nets, progressing in sophistication from the neural <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Neural_Turing_machine" target="_blank" rel="noopener">Turing machine</a> to the <a style="font-family: 'BentonSans', sans-serif;" href="https://en.wikipedia.org/wiki/Differentiable_neural_computer" target="_blank" rel="noopener">differentiable neural computer</a>. Through the inclusion of read-write memory, an attention controller and other features, the latter design in particular has demonstrated an ability to derive algorithms and apply them in new contexts. An age of true “intelligence engineering” may just be beginning.
Q: What will it take to get to the point where AI is not just supporting the human experience, but actually advancing it? 
A: I wish we could lay out a concrete roadmap for AI like the one <a style="font-family: 'BentonSans', sans-serif;" href="http://e-drexler.com/d/07/00/1204TechnologyRoadmap.html" target="_blank" rel="noopener">the nanotechnology community created</a> for their field in 2007. The breadth of AI research, ranging from biology to mathematics to philosophy, would make this extremely difficult, but we can begin by designing our relationship to AI—that is, how we want AI to fit into our lives.
AI shouldn’t be magic. If we don’t know how a model works, we can&#8217;t be sure why it fails, and even when it succeeds we are still excluded. Intelligent products are ones we either like, dislike or distrust.
We know already that we prefer AI systems to <a style="font-family: 'BentonSans', sans-serif;" href="https://designmind.frog.co/2016/02/fostering-smart-software-deep-design-for-deep-learning/" target="_blank" rel="noopener">reveal their mistakes</a> and invite us to correct them. In other words, we want relationships from artificial intelligence, not products. Looking ahead, we can extend this basic etiquette into the principle of transparency, aiming at systems able to both learn and explain themselves simultaneously. Products reflect a designer’s empathy; future AI systems will practice empathy, moment to moment, if we design them to listen and communicate.

Sheldon Pacotti is Senior Solution Architect at frog in Austin. Having studied math and English at MIT and Harvard, Sheldon enjoys cross-disciplinary creative projects. He builds award-winning software, writes screenplays for video games, creates software architectures for businesses and writes about technology.

Illustrations by

Author

How exposing mistakes can improve machine learning.

fostering-smart-software-deep-design-for-deep-learning-640x360

Fostering-Smart-Software-Deep-Design-for-Deep-Learning-640x360

Fostering Smart Software: Deep Design for Deep Learning

As software programs become more like the human mind, they acquire the human mind’s fallibility. Like us, they must learn from mistakes, which they can do only if we teach them. This interdependence is changing our relationship with software and prompting a conversation with these systems that will develop its own rules and etiquette. The way we collectively decide to communicate with our software will define the tone of our daily lives as synthetic intelligence permeates our world.
A common feature of the recent breakthroughs in deep learning is a rate of error. The errors are benign in some domains (speech recognition, captioning a photo) but potentially catastrophic in others (self-driving cars, medicine). What is common about these errors, though, is that they invite us to ask why they happened. We expect an explanation from the underlying system, not so we can fix the problem like a programmer, but so we can reassure ourselves that the mistakes are somehow reasonable. If the mistakes are reasonable, they can be corrected with additional inputs.
This mindset is rare when people approach software today. When a desktop application crashes, we do not ask why. We assume that somewhere in a dull, poorly lit, meager little cubbyhole, some careless coder did something wrong. We do not care what it was, because it was arcane and—teleologically speaking—meaningless. The most we feel is a surge of anger at the company that let one of its errant employees ruin our afternoon. The flaw is in our fellow human.
In contrast, users of speech recognition on desktops are accustomed to inspecting errors via the process of training. You say, “correct ‘the cart was right,’” and a list of other possible matches is displayed. The user can see that the system was considering other possibilities. Often, the user sees the right phrase in the list and simply says “choose three” to change the phrase to “Descartes was right.” The user has both gained confidence in the system’s intelligence and offered input to make the system better.
This mode of interaction is best understood as a conversation. The system says, “I thought you said . . .” and the user replies, “No, actually . . .” This is the aspect missing from many of the new systems making their way into our lives today. How many of us still see items we purchased months ago advertised back to us from our browsers? For me, these are bird-watching binoculars I bought for my wife. How often does our newly sentient economy expect us to buy binoculars? Imagine the accuracy boost that would occur—in both deep and shallow channels—if ads had “like” and “dislike” buttons. By opening a conversation with the recommendation engine, humans would graduate from victim to participant in advertising, while the deep-learning backend would learn more deeply. (Google appears to be experimenting with a similar feedback mechanism, but it is not yet pervasive.)
<blockquote>
<h3>The more appealing vessel for such functionality is a personal “agent,” whose interactions and knowledge we control. But the first step—simply opening up a conversation between user and system—is what we need right now to make the experience better. It is the model we should adopt as the sophistication of software grows.</h3>
</blockquote>
Our current habit is to design only the surface of an experience, striving to make it as simple as possible, which can shut the user out. I am reminded of an experience with a smartwatch that presented a “service not available” message when the watch could not communicate with the phone. It turned out that the web service was not actually down, but since the device was not working or giving me any useful feedback, I was forced to make serendipitous discoveries about how close the devices needed to be, whether to switch from Bluetooth to cellular, and so on, until the deep learning in my brain modified my behavior to accommodate the watch’s limitations. How much easier this would have been if my watch had simply told me what was wrong—like a mature companion—rather than making me guess.
“Deep design” encompasses the full product, from its surface down into its core capabilities. Rather than hiding errors, we should design learning systems that reveal details and ask the user for guidance. “I thought you liked scotch,” the personal assistant of tomorrow might say. “Yes, but not for breakfast!” we will answer.
Of course, training a neural network is not the only form deep design can take. A recent project at frog considered how to show users the possible futures of a complex dataset. Rather than show the user a best guess or a set of likely futures, we decided that the more effective approach—both for visualization and for calculation—was to show all possibilities as probability distributions. This greatly simplified the visual elements that needed to be displayed, while also increasing the information conveyed. We invited the user to view all of the bad guesses alongside the good ones. We were revealing the errors, too, and inviting the user to interpret them, enabling many new interactions, such as checking that a scenario was possible and refining the probabilities via human input.
As research into the human brain continues to reveal the fallacy of our own thinking, leading in some circles to a belief that knowledge is only a momentary viewpoint within “<a href="http://edge.org/response-detail/26733" target="_blank" rel="noopener">a state space of possible world views</a>” and even to suggestions that human choice should join the four humors as an <a href="http://edge.org/response-detail/26792" target="_blank" rel="noopener">antiquated notion</a>, we may find ourselves much more tolerant of fallible machines. In the 1920s, tolerance for automobile fatalities was nearly unthinkable; 60% of fatalities were children, often playing outside their homes. Drivers in all kinds of accidents were charged with manslaughter and paraded through the streets in “<a href="http://www.detroitnews.com/story/news/local/michigan-history/2015/04/26/auto-traffic-history-detroit/26312107/" target="_blank" rel="noopener">safety parades</a>.” We will no doubt face a similarly strident response to self-driving cars in the near future, but just as traffic lights, crosswalks, and the notion of the “accident” helped society balance risk and reward in the automobile era, so will we balance risk and reward as we integrate machine intelligence into all aspects of our lives.
We will succeed by understanding that our machines’ errors make sense in some way, which we can do only by entering into a conversation with them. Via deep learning, some systems have already begun the conversation. They are the first generation of a new order of things being born everywhere from the walls of our houses to the clothes on our bodies. For our own sake, let’s raise them well.

Sheldon Pacotti is Senior Solution Architect at frog in Austin. Having studied math and English at MIT and Harvard, Sheldon enjoys cross-disciplinary creative projects. He builds award-winning software, writes screenplays for video games, creates software architectures for businesses and writes about technology

We should expect another digital revolution, or two, or three.

shapes_bluepink-640x360-1

Shapes_bluepink-640x360 (1)

We Are the Builders of Tech Revolutions. Why Are They Still a Surprise?

In fact, I would like to think we could plot them like stops on a subway line, or – introducing free will – cities on a map.
Nonetheless, I must confess that even being a software engineer, I was puzzled during the ‘90s when I read an article about a company with the odd name of “Amazon” that was building warehouses to replace all of retail. It was like I had woken up to the sound of bulldozers constructing an airport next to my house, and I was a farmer living in 1850. More recently, despite again understanding the constituent technologies, I went through the gee-whiz period of any clueless parent as I discovered just how much TV my kids could watch on their phones.
Let’s agree not to be surprised again.
We are a few years beyond this last revolution, which I characterize as the answer to the question, “Can our software be omnipresent?” But we aren’t done asking questions, or creating revolutions. To me, when I take a long view of computing, I see a logarithmically accelerating sequence of achievements defined by the central problems being solved to realize each achievement. As these problems are solved, surprising shifts in technology and society occur, but these shifts are the answers to a logical sequence of fundamental questions.
concepttimeline_cropped
Technology advances by leaps, resembling the “paradigm shifts” of science as described by Thomas Kuhn—with one exception. In technology, these shifts are of our own making. We have the chance not only to anticipate them but also to bend them to our will.
In science, Kuhn characterizes paradigm shifts as arriving seemingly by surprise after long periods of “normal science.” “[N]ormal science ultimately leads only to the recognition of anomalies and to crisis. And these are terminated, not by deliberation and interpretation, but by a relatively sudden and unstructured event like the gestalt switch.” [1]
To most of us, technological advances feel much like a “gestalt switch,” but they don’t have to. We can take a step back from the headlines of the moment—the big data, the cloud computing, the semantic web, the wearable stocking stuffers of 2014—to ask ourselves what, fundamentally, are we trying to build? What do we need? What would make life easier and more enjoyable?
The preceding chart begins with the era when engineers created a machine that could run any algorithm, then moves to the era when the “operating system” layer was created to house applications, and so on. Though such a timeline can only hope to approximate eras and dates, an acceleration of major paradigm shifts does seem to emerge. (In my defense I stop short of predicting the date of the Singularity.)
My sense, again, is that we are a few years past one of these shifts and are working in a new layer of technology—a “layer of abstraction,” in engineering parlance. That’s how it feels to me, in my role as software architect at frog, where I am called upon repeatedly to grab bits and pieces of functionality from all across the web — social media, Google maps, and cloud storage to name a few. The question of the last era, “Can our software be omnipresent?” has been answered, and we are now turning our attention to the next question: “Can we build new applications from running programs?”
But we aren’t done.
There are other simple questions coming, each difficult to answer but a necessary precursor to the succeeding one. My opinion is that the next question is “Can our applications understand and use each other’s capabilities?” This will abstract the work of today’s engineers into tools for end-users and will provide the necessary foundation for other darlings of the “Web 3.0” vision, such as the autonomous “agent.”
Ultimately, we end up with a “World OS” in which human intention is captured by personal agents able to navigate a universe of honest, well-defined products, services, and—other agents.
Why engage in speculation like this? Because we create these paradigm shifts. We can still shape them, as well as profit greatly by anticipating them. Do you agree that these are the right advances to work toward? Are there open standards that need to be championed in order to enable them?
New standards often destroy or marginalize the killer apps of today, but we should embrace these “gestalt switches.” They are the nature of innovation.
[1] Kuhn, Thomas S. The Structure of Scientific Revolutions: Second Edition, Enlarged. U. of Chicago Press, 1970. Chicago. p. 122.

<div class="divider grid-center ">
<div class="col-12">
<div class="text-content grid-center ">
Sheldon Pacotti is Senior Solution Architect at frog in Austin. Having studied math and English at MIT and Harvard, Sheldon enjoys cross-disciplinary creative projects. He builds award-winning software, writes screenplays for video games, creates software architectures for businesses and writes about technology
</div>
</div>
</div>

Learn about Sheldon Pacotti: view profile, expertise, work, and author contributions on frog, part of Capgemini Invent.

description

index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1

robots

Sheldon Pacotti | frog, part of Capgemini Invent

https://www.frog.co/authors/sheldon-pacotti/

https://frogdesign.nyc3.digitaloceanspaces.com/wp-content/uploads/2020/07/18025455/featured_image_sheldon-pacotti.png

	Global		EN
	France		FR

	Global		EN
	France		FR

Sheldon Pacotti

Featured

It’s Time to Reclaim the Open Office

Articles

Our Studios

Careers